Parameter Estimation Of Direct Sequence Spread Signal In Non-cooperative Communication Environment

The direct sequence spread spectrum(DSSS)signal,which is widely employed in military communications,satellite communications and satellite navigation systems,utilizes the spread spectrum sequence to broaden the signal spectrum so as to meet the condition that the transmission bandwidth is much larger than the bandwidth required to transmit the information.Therefore,the power spectral density of the transmit signal can be reduced and the signal can propagate at low signal to noise ratio,which makes the DSSS signal has a good anti-jamming and anti-intercept characteristic.In non-cooperative communication,these advantages of DSSS communication bring new challenges for signal detection and parameter estimation.It is important to estimate the parameters of the intercepted signal accurately to realize the demodulation and despreading of the signal and finally to capture the original signal.Therefore,the study of the parameter estimation method of the DSSS is an important topic in the field of communication confrontation.In this thesis,we study the parameter estimation and spread code sequence estimation of DSSS signals in non-cooperative situations.The core work and innovation are mainly reflected in the following aspects:1.A novel method for the period estimation of PN code based on correntropy is proposed.By studying the correntropy theory,we find that the correntropy has the characteristic of measuring the local similarity of random process,then the estimation of the period of PN sequence is realized using this characteristic.Firstly,the periodicity implied by the correntropy is analyzed,and the series of expressions proving the periodic existence are theoretically deduced.Secondly,the correntropy of DSSS signal regards to the short code and the long code cases is analyzed respectively.It is proved that the spread spectrum period can be estimated by the correntropy peak interval.Finally,simulation verification experiments are given.2.The delay-multiplied correntropy method for period estimation of the PN sequence is proposed.As the existence of the information code affects the performance of the correntropy method for the estimation of the PN sequence's period,in view of this problem,we introduce the delay multiplication process on the basis of the correntropy method.The correntropy method and its improved algorithm are discussed and compared with the cepstrum method,the spectrum reprocessing method based on power spectrum,and the fluctuation correlation method based on autocorrelation.The estimation performance of the proposed methods based on correntropy is much better.3.An improved sliding correlation method is proposed in order to estimate the starting position for synchronization of the PN sequence.The traditional sliding correlation method is to segment the received signal by using a sliding window with the length of a period of the PN code,then the starting symbol position of synchronization of the PN sequence is determined according to the maximum mean value of the correlation function of each segmented sequence under different starting points of the sliding window.Considering the problem that the existence of the information code affects the performance of the algorithm,an improvement is proposed as following: before calculating the mean value of each correlation value,we first get the absolute value of it,and the correlation values are non-negative,so the average operation becomes a process which can accumulate and increase the peak value,thereby the algorithm performance is improved.4.Estimation of the starting position for synchronization of the PN sequence by correntropy method is provided.Sliding the starting point of the DSSS signal and combining with the segmentation processing by the period of PN sequence,the correntropy value of each sequence is calculated and filtered based on the local similarity of the correntropy.For the different starting symbol positions,the maximum value is picked out between calculation results at different starting points.At last,the estimation performance of this method is verified by simulation experiments.5.A method of starting position estimation for synchronization of the PN sequence based on average inner product is presented in this thesis.The method first slides the starting point,and the signal sequence is segmented by the window with the length of the PN sequence's period to construct the signal matrix.By searching for the biggest average value of the absolute product for any two sub-sequences,the synchronization point of the PN sequence can be estimated.6.Aiming at the problem of high computational complexity of the existing PN sequence estimation methods,a new method based on correlation decision function is proposed.It is assumed that the synchronization position of the PN sequence is known.By using the window with length of the period of PN sequence,the signal,sampled at the rate of the PN sequence is segmented into several subsequences.Then a signal matrix is constructed,with the information symbols being the row number and period length of the PN sequence being the column number.Using the correlative decision function to get the dot product between each column and the first column of the matrix,then we can judge the symbolic relationship between all other symbols and the first symbol of the PN sequence.That is,the symbol of other PN sequence symbols can be determined by its first symbol,thereby the complete PN sequence can be recovered.Both the theoretical derivation and the simulation experiments show the effectiveness of the proposed method,and the computational complexity of the algorithm is greatly reduced compared with the common PN sequence estimation algorithm based on EVD and TCF.